Multi-parameter intelligent optimization method and system for wide-temperature-range lithium battery electrolyte

Abstract

The invention relates to the technical field of lithium batteries, and discloses a multi-parameter intelligent optimization method and system for a wide-temperature-range lithium battery electrolyte. The method comprises the following steps: acquiring electrolyte parameters, wherein the electrolyte parameters comprise formula parameters and process parameters; presetting constraint conditions, and constructing a data driving model according to the formula parameters, the process parameters and the constraint conditions; performing process simulation on each formula parameter by using the same process parameter to obtain an optimal formula parameter, and performing process simulation on each process parameter by using the same formula parameter to obtain an optimal process parameter; by comprehensively considering the mutual influence relationship between the formula parameters and the process parameters of the electrolyte and randomly simulating and calculating the score after the formula parameters and the process parameters are combined, collaborative optimization among multiple parameters is realized, and the comprehensive performance of the electrolyte in a wide temperature range is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of lithium batteries, and more specifically, to a multi-parameter intelligent optimization method and system for wide-temperature-range lithium battery electrolytes. Background Art

[0002] As an efficient and portable energy storage device, lithium batteries have been widely used in many fields such as electric vehicles, portable electronic devices, and energy storage power stations; however, traditional lithium batteries have certain limitations in temperature adaptability, which greatly affects their performance and safety;

[0003] In practical applications, lithium batteries often need to work in different temperature environments; for example, electric vehicles need to maintain good performance in extremely cold winters and hot summers; the electrolytes that affect the performance of lithium batteries have significantly reduced ionic conductivity in low-temperature environments, resulting in poor charge and discharge performance of the batteries and accelerated capacity decay; moreover, low temperature may also trigger the growth of lithium dendrites, increasing the risk of battery short circuit; in high-temperature environments, the electrolyte is prone to decomposition reactions, generating gases, increasing the internal pressure of the battery, and even possibly triggering thermal runaway.

[0004] To ensure the stable operation of the electrolyte in a wide temperature range, it is necessary to determine multiple parameters that affect the performance of lithium battery electrolytes, including the selection of solvents, the types and contents of additives, the concentration of electrolyte salts, preparation process conditions, etc.; traditional electrolyte optimization methods mainly rely on a large number of experiments and experience accumulation, which is not only time-consuming and laborious, but also difficult to comprehensively consider the complex interaction between multiple parameters; for example, the Chinese patent application with the authorization announcement number CN117410568B discloses a wide-temperature-range high-voltage lithium battery electrolyte and its preparation method; for another example, the patent application with the publication number CN106785024A discloses a wide-temperature-range and long-life lithium iron phosphate battery electrolyte and its preparation method; therefore, there is an urgent need for an intelligent process parameter optimization method;

[0005] In view of this, the present invention proposes a multi-parameter intelligent optimization method and system for wide-temperature-range lithium battery electrolytes to solve the above problems. Summary of the Invention

[0006] In order to overcome the above-mentioned defects of the prior art, the present invention provides the following technical solutions: A multi-parameter intelligent optimization method for wide-temperature-range lithium battery electrolytes, comprising:

[0007] Obtain electrolyte parameters, where the electrolyte parameters include formulation parameters and process parameters;

[0008] Preset constraint conditions, and construct a data-driven model according to the formulation parameters, process parameters, and constraint conditions;

[0009] Perform process simulations on each formulation parameter with the same process parameters to obtain the optimal formulation parameters, and perform process simulations on each process parameter with the same formulation parameters to obtain the optimal process parameters;

[0010] Randomly combine the optimal process parameters corresponding to each formulation parameter and the optimal formulation parameters corresponding to each process parameter into k electrolyte parameters, and perform optimization calculations on the k electrolyte parameters to obtain the optimal electrolyte parameters as the intelligent optimization parameters.

[0011] Preferably, both the formulation parameters and the process parameters are composed of a combination of multiple sub-parameters. The formulation parameters are marked as x, x = [x 1 , x 2 … x a , where x a is the a-th sub-parameter in the electrolyte parameters; the process parameters are marked as y, y = [y 1 , y 2 … y b , where y b is the b-th sub-parameter in the process parameters.

[0012] Preferably, the constraint conditions include the preset electrolyte simulation temperature C, the lowest simulation temperature C MIN , the highest simulation temperature C MAX , the conductivity peak σ(T) MAX and the viscosity valley value ND MIN , where C MIN < C < C MAX , C MAX > 0, C MIN < 0;

[0013] The data-driven model is a mathematical and physical model for numerically describing and calculating electrolyte parameters.

[0014] Preferably, the method for obtaining the optimal formulation parameters by performing process simulations on each formulation parameter with the same process parameters includes:

[0015] Obtain m' simulation results after performing process simulations on the same process parameters under m groups of different formulation parameters. The simulation results include conductivity, ion diffusion coefficient, and electrochemical window. The simulation results are used as performance indicators, m' = m, and the simulation results correspond one-to-one with the formulation parameters;

[0016] Input the performance indicators corresponding to each formulation parameter into the trained optimization analysis model, predict the judgment labels corresponding to each formulation parameter, obtain the corresponding judgment results according to the predicted judgment labels, and determine the optimization direction of the performance indicators corresponding to the formulation parameters. The optimization direction is the tendency towards the peak or the valley of the performance indicators during the optimization process;

[0017] According to the judgment labels corresponding to each formulation parameter, n' simulation results are selected from the m' simulation results corresponding to each formulation parameter, and n sets of corresponding formulation parameters are obtained according to the n' simulation results and marked as excellent formulation parameters; where n = n', m > n > 0.

[0018] Preferably, the method for obtaining the optimal process parameters by performing process simulations on each process parameter with the same formulation parameter includes:

[0019] Obtain g' simulation results after performing process simulations on the same formulation parameter under g sets of different process parameters. The simulation results include conductivity, ion diffusion coefficient, and electrochemical window. The simulation results are used as performance indicators, g' = g, and the simulation results correspond one-to-one with the process parameters;

[0020] Input the performance indicators corresponding to each process parameter into the trained optimization analysis model to predict the judgment label corresponding to each process parameter, obtain the corresponding judgment result according to the predicted judgment label, and judge the optimization direction of the performance indicator corresponding to the process parameter. The optimization direction is the tendency peak or tendency valley of the performance indicator during the optimization process;

[0021] According to the judgment label corresponding to each process parameter, j' simulation results are selected from the g' simulation results corresponding to each process parameter, and j sets of corresponding process parameters are obtained according to the j' simulation results and marked as excellent process parameters; where j = j', g > j > 0.

[0022] Preferably, the training process of the optimization analysis model includes:

[0023] Pre-set corresponding judgment results for multiple groups of performance indicators. The judgment results include tendency peak and tendency valley, and different digital labels are set for both the tendency peak and the tendency valley;

[0024] Mark the digital label of the judgment result as the judgment label, and convert the performance indicator and the corresponding judgment label into a corresponding set of feature vectors;

[0025] Take each set of feature vectors as the input of the optimization analysis model. The optimization analysis model outputs a set of predicted judgment labels corresponding to each set of performance indicators, takes the actual judgment label corresponding to each set of performance indicators as the prediction target. The actual judgment label is the digital label of the pre-set judgment result corresponding to the performance indicator; take minimizing the sum of the prediction errors of all performance indicators as the training target; train the optimization analysis model until the sum of the prediction errors reaches convergence and then stop training; the optimization analysis model is a deep neural network model.

[0026] Preferably, the optimal process parameters corresponding to each formulation parameter and the optimal formulation parameters corresponding to each process parameter are randomly combined into k electrolyte parameters, where k = n × j, and the electrolyte parameters are denoted as p, and p = (x, y);

[0027] Simulate the k electrolyte parameters, and obtain the optimal electrolyte parameters as the intelligent optimization parameters.

[0028] Preferably, the method for obtaining the optimal electrolyte parameters includes:

[0029] S1: Preset the initial temperature, the cooling coefficient δ, and the maximum number of iterations And set the initial temperature = the highest simulation temperature C MAX ;

[0030] S2: Randomly set a feasible solution S, where the feasible solution S is the electrolyte parameter p, the range of the electrolyte parameters is the k electrolyte parameters, and the range of the electrolyte parameters is the range of the feasible solution S;

[0031] S3: Determine the fitness function;

[0032] The expression of the fitness function is: f(x, y) = DJ, where f(x, y) is the fitness and DJ is the comprehensive score of the electrolyte parameters;

[0033] S4: Calculate the fitness f(x, y) corresponding to the feasible solution S; taking the feasible solution S as the current point, perform random perturbation within the neighborhood of the current point to obtain a new feasible solution S′, and calculate the fitness f(x, y)′ corresponding to the new feasible solution S′;

[0034] S5: Calculate the fitness difference f(x, y)″, and the expression of the fitness difference f(x, y)″ is f(x, y)″ = f(x, y)′ - f(x, y);

[0035] If the fitness difference f(x, y)″ > 0, then let S = S′, that is, assign the value of the new feasible solution S′ to the feasible solution S; if the fitness difference f(x, y)″ ≤ 0, then calculate the probability k′, and let S = S′ according to the probability k′; the expression of the probability k′ is: In the formula, e is the natural constant;

[0036] S6: Loop S4 - S5 until the number of loops reaches the maximum number of iterations When it reaches, the loop ends and enters S7;

[0037] S7: Let the current temperature C DQ = C × δ, that is, cool down the current temperature in S1, and assign the cooled value to the current temperature; let the maximum number of iterations Assign the value of the reduced maximum number of iterations to the maximum number of iterations; if the reduced maximum number of iterations is not an integer, round up the reduced maximum number of iterations to make it an integer.

[0038] S8: Loop through S4 - S7 until the current temperature C DQ < C min When the loop ends, obtain the electrolyte parameters corresponding to the feasible solution S and mark them as the optimal electrolyte parameters.

[0039] Preferably, the calculation method for the comprehensive score of the electrolyte parameters includes:

[0040] DJ = ω 1 × f 1 (x, y) + ω 2 × f 2 (x, y) + ω 3 × f 3 (x, y);

[0041] In the formula, f 1 (x, y) is the conductivity, f 2 (x, y) is the ion diffusion coefficient, f 3 (x, y) is the electrochemical window, ω 1 , ω 2 and ω 3 are the weights of each objective, x is the formulation parameter, and y is the process parameter.

[0042] A multi-parameter intelligent optimization system for wide-temperature lithium battery electrolytes, used to implement the above-mentioned multi-parameter intelligent optimization method for wide-temperature lithium battery electrolytes, includes:

[0043] A data acquisition module, used to collect electrolyte parameters, where the electrolyte parameters include formulation parameters and process parameters, and used to collect the performance data of the electrolyte at different temperatures;

[0044] A data-driven module, used to set constraint conditions and build a data-driven model based on the formulation parameters, process parameters, and constraint conditions;

[0045] A process simulation module, which performs process simulation based on the formulation parameters and process parameters to obtain the optimal formulation parameters and the optimal process parameters;

[0046] An optimization calculation module: Randomly combine the optimal process parameters corresponding to each formulation parameter and the optimal formulation parameters corresponding to each process parameter into k electrolyte parameters, and perform optimization calculations on the k electrolyte parameters to obtain the optimal electrolyte parameters as the intelligent optimization parameters.

[0047] Technical effects and advantages of the multi-parameter intelligent optimization method and system for wide-temperature-range lithium battery electrolytes of the present invention:

[0048] 1. By comprehensively considering the mutual influence relationship between the formulation parameters and process parameters of the electrolyte, respectively simulating the excellent process parameters of different process parameters under the same formulation parameters, as well as the excellent formulation parameters of different formulation parameters under the same process parameters, and randomly simulating and calculating the scores after the combination of the two, the collaborative optimization between multiple parameters is achieved, and the comprehensive performance of the electrolyte within a wide temperature range is improved.

[0049] 2. Through random simulation and intelligent optimization methods, quickly converge to the global optimal solution, and fully consider the balance between multiple objectives to ensure that performance indicators such as the conductivity, ion diffusion coefficient, and electrochemical window of the electrolyte reach the optimal. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a flowchart of the multi-parameter intelligent optimization method for wide-temperature-range lithium battery electrolytes according to Embodiment 1 of the present invention;

[0051] Figure 2 It is an improved flowchart of the multi-parameter intelligent optimization method for wide-temperature-range lithium battery electrolytes according to Embodiment 1 of the present invention;

[0052] Figure 3 It is a structural diagram of the multi-parameter intelligent optimization system for wide-temperature-range lithium battery electrolytes according to Embodiment 2 of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0053] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0054] Embodiment 1

[0055] Please refer to Figure 1 and Figure 2 As shown, this embodiment discloses a multi-parameter intelligent optimization method for wide-temperature-range lithium battery electrolytes, including:

[0056] Obtain electrolyte parameters, where the electrolyte parameters include formulation parameters and process parameters; both the formulation parameters and the process parameters are composed of a variety of sub-parameters. The formulation parameters are marked as x, x = [x 1 , x 2 … x a , x a is the a-th sub-parameter in the electrolyte parameters; the process parameters are marked as y, y = [y1 , y 2 …y b , y b is the b-th sub-parameter among the process parameters; the sub-parameters in the formulation parameters include but are not limited to solvent ratio, electrolyte concentration, additive type, etc.; the sub-parameters in the process parameters include but are not limited to stirring speed, mixing temperature, mixing time, etc., which are relevant parameters affecting the performance data of the finished electrolyte under the same electrolyte formulation, and are not specifically defined here. The formulation parameters in this embodiment are obtained according to the electrolyte preparation information manual or obtained by combining the actual experience of those skilled in the art;

[0057] Exemplarily, the solvents include ethylene carbonate (EC), dimethyl carbonate (DMC), ethyl methyl carbonate (EMC), etc. By combining solvents in different ratios, the viscosity, conductivity, and low-temperature performance of the electrolyte can be affected. For example, the ratio of EC:DMC:EMC can be set to 3:5:2 or 2:6:2 for simulation to determine the optimal formulation parameters; again, the concentration range of the lithium salt is usually between 0.5 M and 1.5 M. A high concentration increases the ion transference number but may increase the viscosity. Therefore, the optimal balance value needs to be found. Additive types: include SEI film-forming additives (such as VC), high-temperature inhibitors (such as LiBOB), flame retardants (such as TMSP), etc. The addition ratio of each additive can be adjusted during optimization; for example, the flame retardant can improve the safety of the electrolyte;

[0058] The process parameters are set as follows: the stirring speed is 200 - 1000 rpm. If the stirring speed is too low, it will cause uneven mixing, and if it is too high, it will cause bubble generation; the mixing temperature range is 20 - 60 °C. If the temperature is too high, it may cause solvent volatilization, and if it is too low, it may increase the mixing viscosity; the mixing time is 10 - 60 minutes to ensure the uniformity of the formulation. Therefore, even with the same formulation parameters, different process parameters will affect the final electrolyte performance.

[0059] Preset constraint conditions, and construct a data-driven model based on the formulation parameters, process parameters, and constraint conditions;

[0060] The data-driven model is a mathematical and physical model for numerically describing and calculating the electrolyte formulation; the simulation model in this embodiment is established through the Arrhenius equation and simulation software. Exemplarily, the Arrhenius equation describes the effect of temperature on the reaction rate and is used to simulate the performance changes of the electrolyte at different temperatures. The specific expression is: In the formula, dk is the rate constant (such as the change in conductivity), is the pre-factor, and the specific value is obtained from the raw material experiments of the formulation parameters or data fitting. E ais the activation energy, which determines the sensitivity of the reaction rate to temperature and is obtained through historical electrolyte processing data or experiments. R is the gas constant, and T is the temperature in Kelvin (K). For example, by fitting experimental data, the variation law of conductivity within the range of -40°C to 80°C for different formulations can be obtained. Simulation software, such as multi-physics simulation tools like COMSOL and ANSYS, is used to establish the mathematical and physical model of the electrolyte to simulate performance indicators such as viscosity, conductivity, and ion diffusion.

[0061] The constraint conditions include the preset simulated temperature C of the electrolyte, the lowest simulated temperature C MIN and the highest simulated temperature C. MAX The conductivity peak value σ(T), MAX the viscosity valley value ND, MIN the ion diffusion coefficient, and the electrochemical stability window. Among them, C MIN < C < C, MAX where C MAX > 0, C MIN < 0. The purpose is to simulate the performance of electrolyte parameters under wide temperature range conditions, such as simulation between -40°C and 80°C. For example, the simulation temperature range is set from -40°C to 80°C, and the simulation is carried out in steps of 5°C. The focus is on the low-temperature conductivity at -40°C and the high-temperature stability at 80°C. The conductivity peak value constraint, such as the conductivity at room temperature of 25°C needs to be greater than 8.0 mS / cm. The conductivity peak value limit is set to ≥ 8.0 mS / cm. The viscosity valley value constraint, such as the viscosity at -40°C is less than 10 mPa·s. The ion diffusion coefficient needs to reach a certain threshold, such as 1×10 -6 cm 2 / s. The electrochemical stability window needs to be greater than 4.5 V to ensure the cycle stability within the wide temperature range. The specific constraint conditions are determined by those skilled in the art according to the actual situation and are not specifically limited here.

[0062] Process simulation is carried out for each formulation parameter with the same process parameters to obtain the optimal formulation parameters, and process simulation is carried out for each process parameter with the same formulation parameters to obtain the optimal process parameters.

[0063] Specifically, the method for obtaining the optimal formulation parameters by carrying out process simulation for each formulation parameter with the same process parameters includes:

[0064] Obtain m' simulation results after process simulation under m groups of different formulation parameters with the same process parameters. The simulation results include but are not limited to conductivity, viscosity, ion diffusion coefficient, and electrochemical stability window, which are performance indicators. m' = m, and the simulation results correspond one-to-one with the formulation parameters.

[0065] Input the performance indicators corresponding to each formulation parameter into the trained optimization analysis model to predict the judgment labels corresponding to each formulation parameter; obtain the corresponding judgment results according to the predicted judgment labels, and judge the optimization direction of the performance indicators corresponding to the formulation parameters. The optimization direction is the tendency towards the peak or the valley value of the performance indicator during the optimization process;

[0066] According to the judgment labels corresponding to each formulation parameter, screen out n' simulation results from the m' simulation results corresponding to each formulation parameter, obtain the corresponding n groups of formulation parameters according to the n' simulation results, and label them as excellent formulation parameters; where n = n', m > n > 0;

[0067] The method for obtaining the optimal process parameters by performing process simulations on each process parameter with the same formulation parameter includes:

[0068] Obtain g' simulation results after performing process simulations on the same formulation parameter under g groups of different process parameters. The simulation results are used as performance indicators, g' = g, and the simulation results correspond one-to-one with the process parameters;

[0069] Input the performance indicators corresponding to each process parameter into the trained optimization analysis model to predict the judgment labels corresponding to each process parameter; obtain the corresponding judgment results according to the predicted judgment labels, and judge the optimization direction of the performance indicators corresponding to the formulation parameters. The optimization direction is the tendency towards the peak or the valley value of the performance indicator during the optimization process;

[0070] According to the judgment labels corresponding to each process parameter, screen out j' simulation results from the g' simulation results corresponding to each process parameter, obtain the corresponding j groups of process parameters according to the j' simulation results, and label them as excellent process parameters; where j = j', g > j > 0;

[0071] The training process of the optimization analysis model includes:

[0072] Pre-set corresponding judgment results for multiple groups of performance indicators. The judgment results include the tendency towards the peak and the tendency towards the valley value, and different numerical labels are set for both the tendency towards the peak and the tendency towards the valley value. Exemplarily, set the numerical label for the tendency towards the peak as 0, and set the numerical label for the tendency towards the valley value as 1; the tendency towards the peak indicates that the performance indicator should be optimized towards the maximum value, that is, the larger the performance indicator, the better. For example, the higher the conductivity, the better; the tendency towards the valley value indicates that the performance indicator should be optimized towards the minimum value, that is, the smaller the performance indicator, the better. For example, the lower the viscosity at low temperature, the better. The judgment results corresponding to the performance indicators are collected by those skilled in the art for each performance indicator corresponding to each formulation parameter during the preparation of the electrolyte. Those skilled in the art judge the judgment labels corresponding to multiple different performance indicators in sequence according to experience, and set the corresponding judgment results for multiple different performance indicators in sequence;

[0073] Mark the digital label of the judgment result as the judgment label, and convert the performance index and the corresponding judgment label into a corresponding set of feature vectors;

[0074] Use each set of feature vectors as the input of the optimization analysis model. The optimization analysis model takes a set of predicted judgment labels corresponding to each set of performance indicators as the output, and the actual judgment label corresponding to each set of performance indicators as the prediction target. The actual judgment label is the digital label of the judgment result corresponding to the performance indicator set in advance; take minimizing the sum of the prediction errors of all performance indicators as the training target; where the calculation formula of the prediction error is Z g =(α g -μ g ) 2 , where Z g is the prediction error, g is the group number of the feature vectors corresponding to the performance indicator, α g is the predicted judgment label corresponding to the g-th group of performance indicators, and μ g is the actual judgment label corresponding to the g-th group of performance indicators; train the optimization analysis model until the sum of the prediction errors converges and then stop training;

[0075] The above optimization analysis model is specifically a deep neural network model;

[0076] Exemplarily, in the formulation parameters, the electrolyte concentration, such as the type and concentration of lithium salt, will directly affect the ionic conductivity of the electrolyte; the solvent type: organic solvents (such as carbonates) or solid electrolytes, will affect the low-temperature performance and high-temperature stability; additives: such as ionic liquids, stabilizers, etc., can optimize the battery performance or improve the temperature adaptability; by optimizing the ratio of the electrolyte formulation, the performance of the electrolyte can be changed; for example, viscosity affects the ion transport in the battery, and too high or too low viscosity will affect the battery performance; conductivity: ensure good conductivity can be maintained under different temperature conditions, so the conductivity tends to peak;

[0077] Sort the m' simulation results with the same process parameters but different formulation parameters from large to small, and select the formulation parameters corresponding to the top n' simulation results as the excellent formulation parameters.

[0078] Randomly combine the best process parameters corresponding to each formulation parameter and the best formulation parameters corresponding to each process parameter into k electrolyte parameters, and perform optimization calculations on the k electrolyte parameters to obtain the optimal electrolyte parameters as the intelligent optimization parameters; specifically, randomly combine the best process parameters corresponding to each formulation parameter and the best formulation parameters corresponding to each process parameter into k electrolyte parameters, k = n×j, the electrolyte parameter is marked as p, p=(x, y), where, x = [x 1 ,x 2 …x a , x ais the a-th sub-parameter in the formulation parameter x, such as the solvent ratio, electrolyte concentration, and additive concentration, y = [y 1 , y 2 … y b , y a is the b-th sub-parameter in the process parameter y, such as the stirring speed, temperature, and mixing time; the purpose is to first screen out the optimal formulation parameters and the optimal process parameters, and then simulate by randomly combining the optimal formulation parameters and the optimal process as the electrolyte parameters, optimize and calculate the k electrolyte parameters, obtain the optimal electrolyte parameters as the intelligent optimization parameters, reduce the calculation amount, improve the efficiency of multi-parameter intelligent optimization, and considering the mutual influence between the process parameters and the formulation parameters, improve the accuracy of the electrolyte parameters.

[0079] The method for obtaining the optimal electrolyte parameters includes:

[0080] S1: Preset the initial temperature, the lowest temperature C MIN , the cooling coefficient δ, and the maximum number of iterations and let the current temperature C = the initial temperature = the highest simulation temperature C MAX ;

[0081] S2: Randomly set a feasible solution S, and the feasible solution S is the electrolyte parameter p. The range of the electrolyte parameter is k electrolyte parameters, and the range of the electrolyte parameter is the range of the feasible solution S;

[0082] S3: Determine the fitness function;

[0083] The expression of the fitness function is: f(x, y) = DJ, where f(x, y) is the fitness and DJ is the comprehensive score of the electrolyte parameters;

[0084] The calculation method of the comprehensive score of the electrolyte parameters includes:

[0085] DJ = ω 1 × f 1 (x, y) + ω 2 × f 2 (x, y) + … + ω i × f i (x, y);

[0086] In the formula, f i (x, y) is the i-th type of performance index after the electrolyte parameter simulation, f 1 (x, y) is the conductivity, f 2 (x, y) is the ion diffusion coefficient, f 3 (x, y) is the electrochemical window, x is the formulation parameter, y is the process parameter, ω 1 , ω 2 , …, ω iis the weight of each target, satisfying ω 1 , ω 2 , …, ω i The sum is 1. In this implementation, ω 1 , ω 2 , …, ω i The specific values can be set according to the actual situation. The weight coefficient reflects the influence degree of each performance index on the comprehensive score of electrolyte parameters. Those skilled in the art can preset the corresponding weight coefficient according to the actual influence degree of each performance index on the comprehensive score of electrolyte parameters, so as to accurately evaluate the performance of electrolyte parameters;

[0087] It should be noted that the performance index is the influencing parameter of the comprehensive score of electrolyte parameters. For example, the greater the conductivity, the greater the comprehensive score of electrolyte parameters, and vice versa; Since the comprehensive score of electrolyte parameters is only used to reflect the performance of electrolyte parameters, the calculation of the battery comprehensive score is a dimensionless calculation;

[0088] S4: Calculate the fitness f(x, y) corresponding to the feasible solution S; Take the feasible solution S as the current point, perform random perturbation in the neighborhood of the current point to obtain a new feasible solution S′, and calculate the fitness f(x, y)′ corresponding to the new feasible solution S′;

[0089] S5: Calculate the fitness difference f(x, y)″, and the expression of the fitness difference f(x, y)″ is f(x, y)″ = f(x, y)′ - f(x, y);

[0090] If the fitness difference f(x, y)″ > 0, then let S = S′, that is, assign the value of the new feasible solution S′ to the feasible solution S; If the fitness difference f(x, y)″ ≤ 0, then calculate the probability k′, and let S = S′ according to the probability k′; The expression of the probability k′ is: where e is the natural constant;

[0091] S6: Loop S4 to S5 until the number of loops reaches the maximum number of iterations At this time, the loop ends and enters S7;

[0092] S7: Let the current temperature C DQ = C × δ, that is, cool down the current temperature in S1 and assign the cooled value to the current temperature; Let the maximum number of iterations That is, assign the reduced value of the maximum number of iterations to the maximum number of iterations; If the reduced maximum number of iterations is not an integer, then round up the reduced maximum number of iterations to make the reduced maximum number of iterations an integer;

[0093] S8: Loop S4 to S7 until the current temperature C DQ < C minWhen the loop ends, the electrolyte parameters corresponding to the feasible solution S are obtained and marked as the optimal electrolyte parameters, that is, under the wide-temperature conditions, the electrolyte parameters are gradually screened, and finally the electrolyte parameters with the best performance indicators are selected between the highest temperature and the lowest temperature;

[0094] It should be noted that the initial temperature C MAX , the lowest temperature C MIN , the cooling coefficient δ, and the maximum number of iterations are used as preset parameters. The preset parameters are collected by those skilled in the art from Q analysis sets during the historical electrolyte preparation. Each analysis set includes k electrolyte parameters. For the same analysis set, multiple different groups of preset parameters are sequentially preset, and the simulated annealing algorithm is used to obtain the electrolyte parameters in turn. The fitness corresponding to multiple electrolyte parameters is calculated, the electrolyte parameters corresponding to the maximum fitness are obtained and marked as the maximum electrolyte parameters; the preset parameters corresponding to the maximum electrolyte parameters are used as the preset parameters corresponding to the analysis set, and so on to obtain the preset parameters corresponding to Q analysis sets. The mean values of multiple preset parameters (i.e., the mean value of the initial temperature, the mean value of the lowest temperature, the mean value of the cooling coefficient, and the mean value of the maximum number of iterations) are used as the preset initial temperature, the lowest temperature C MIN , the cooling coefficient δ, and the maximum number of iterations and let the current temperature C = the initial temperature = the highest simulated temperature C MAX ;

[0095] In this embodiment, by comprehensively considering the mutual influence relationship between the formulation parameters and process parameters of the electrolyte, excellent process parameters under different process parameters with the same formulation parameters and excellent formulation parameters under different formulation parameters with the same process parameters are respectively simulated, and the scores after the combination of the two are randomly simulated and calculated, realizing the collaborative optimization between multiple parameters and improving the comprehensive performance of the electrolyte within a wide temperature range; through random simulation and intelligent optimization methods, it quickly converges to the global optimal solution, and fully considers the balance between multiple objectives to ensure that performance indicators such as the conductivity, ion diffusion coefficient, and electrochemical window of the electrolyte reach the optimum.

[0096] Embodiment 2

[0097] Please refer to Figure 3 as shown, for the multi-parameter intelligent optimization system of the wide-temperature lithium battery electrolyte, which is used to implement the above-mentioned multi-parameter intelligent optimization method for the wide-temperature lithium battery electrolyte, including a data acquisition module, a data-driven module, a process simulation module, and an optimization calculation module, where each module is connected by wire and / or wirelessly:

[0098] A data acquisition module for collecting electrolyte parameters, where the electrolyte parameters include formulation parameters and process parameters, and for collecting the performance data of the electrolyte at different temperatures; in this embodiment, the data acquisition module is obtained by using a high performance liquid chromatograph (HPLC), a gas chromatograph (GC) and an ICP-MS, and no specific limitation is made here.

[0099] It is worth mentioning that the process parameters are provided by a blender (with speed and time control functions), a temperature-controlled mixing device, etc.; the performance data is obtained in an experimental environment, and performance testing equipment such as a conductivity meter is used to measure the conductivity in a wide temperature range (-40 °C to 80 °C); a rotational viscometer is used to measure the viscosity of the electrolyte at different temperatures; an electrochemical workstation is used to test the electrochemical window and the ion diffusion coefficient.

[0100] A data-driven module for setting constraint conditions and constructing a data-driven model based on the formulation parameters, process parameters and constraint conditions: setting the boundary conditions of the performance indexes according to the design objectives, for example: conductivity ≥ 8.0 mS / cm; viscosity ≤ mPa·s; ion diffusion coefficient ≥ 1×10 -6 cm 2 / s; the tools used to construct the data-driven model include modeling tools and simulation tools, etc. The modeling tools such as Python use scikit-learn, TensorFlow or PyTorch to implement the training of the data-driven model; the simulation tools such as COMSOL Multiphysics; a mathematical and physical model of temperature and performance is established based on the Arrhenius equation; MATLAB is used for the preliminary simulation of parameter optimization.

[0101] A process simulation module for performing process simulation based on the formulation parameters and process parameters to obtain the optimal formulation parameters and the optimal process parameters;

[0102] An optimization calculation module: used to randomly combine the optimal process parameters corresponding to each formulation parameter and the optimal formulation parameters corresponding to each process parameter into k electrolyte parameters, and perform optimization calculation on the k electrolyte parameters to obtain the optimal electrolyte parameters as the intelligent optimization parameters.

[0103] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the above-mentioned division of each functional unit and module is used as an example. In actual applications, the above-mentioned functions can be assigned to different functional units and modules according to needs, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiment can be integrated into a processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit. In addition, the specific names of each functional unit and module are only for the convenience of mutual distinction and do not limit the protection scope of this application. The specific working processes of the units and modules in the above system can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0104] In the above embodiments, the descriptions of each embodiment have their own emphases. For the parts not detailed or recorded in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0105] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or by a combination of computer software and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0106] In the embodiments provided in this application, it should be understood that the disclosed device / terminal device and method can be implemented in other ways. For example, the device / terminal device embodiments described above are only illustrative. For example, the division of the module or unit is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection to each other can be through some interfaces. The indirect coupling or communication connection of the device or unit can be in electrical, mechanical or other forms.

[0107] In addition, each functional unit in the various embodiments of this application can be integrated into a processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit.

[0108] The unit described as a separation component may or may not be physically separated. The component shown as a unit may or may not be a physical unit, that is, it may be located in one place or may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0109] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, to implement all or part of the processes in the above-mentioned embodiment methods of this application, it can also be completed by instructing relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above-mentioned various method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, RandomAccessMemory), electrical carrier signal, telecommunication signal, and software distribution medium, etc.

[0110] As mentioned above, the above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

[0111] Finally: The above is only the preferred embodiment of the present invention and is not used to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A multi-parameter intelligent optimization method for wide temperature range lithium battery electrolyte, characterized in that: include: Obtaining electrolyte parameters, which include formula parameters and process parameters; Preset constraints and build data-driven models based on recipe parameters, process parameters and constraints; The same process parameters are used to simulate the process for each recipe parameter to obtain the optimal recipe parameters. The same process parameters are used to simulate the process for each recipe parameter to obtain the optimal process parameters. The optimal process parameters corresponding to each formula parameter and the optimal formula parameters corresponding to each process parameter are randomly combined into k electrolyte parameters, the k electrolyte parameters are optimized and calculated, and the optimal electrolyte parameters are obtained as intelligent optimization parameters.

2. The multi-parameter intelligent optimization method for wide temperature range lithium battery electrolyte according to claim 1, characterized in that: The formula parameters and process parameters are composed of a combination of multiple sub-parameters. The formula parameter is marked as x, where x = [x1, x2…x a ], x a is the ath seed parameter in the electrolyte parameter; the process parameter is marked as y, y=[y1,y2…y b ],y b It is the bth seed parameter among the process parameters.

3. The multi-parameter intelligent optimization method for wide temperature range lithium battery electrolyte according to claim 2, characterized in that: The constraints include a preset electrolyte simulation temperature C, a minimum simulation temperature C MIN , Maximum simulation temperature C MAX 、Conductivity peak σ(T) MAX And the viscosity valley value ND MIN , where C MIN <C<C MAX , C MAX >0,C MIN <0; The data-driven model is a mathematical and physical model for numerically describing and calculating electrolyte parameters.

4. The multi-parameter intelligent optimization method for wide temperature range lithium battery electrolyte according to claim 3, characterized in that: Each recipe parameter uses the same process parameters to perform process simulation. The methods for obtaining the best recipe parameters include: Obtain m′ simulation results after process simulation under m groups of different recipe parameters with the same process parameters. The simulation results include conductivity, ion diffusion coefficient and electrochemical window. The simulation results are used as performance indicators, m′=m, and the simulation results correspond to the recipe parameters one by one. Input the performance index corresponding to each recipe parameter into the trained optimization analysis model, predict the judgment label corresponding to each recipe parameter, obtain the corresponding judgment result according to the predicted judgment label, and judge the optimization direction of the performance index corresponding to the recipe parameter. The optimization direction is the peak value or valley value of the performance index during the optimization process. According to the judgment label corresponding to each recipe parameter, n′ simulation results are screened out from the m′ simulation results corresponding to each recipe parameter, and the corresponding n groups of recipe parameters are obtained according to the n′ simulation results, and marked as excellent recipe parameters; wherein n=n′, m>n>0.

5. The multi-parameter intelligent optimization method for wide temperature range lithium battery electrolyte according to claim 4, characterized in that: The same recipe parameters are used to simulate each process parameter. The methods for obtaining the optimal process parameters include: Obtain g′ simulation results after process simulation under g groups of different process parameters with the same recipe parameters. The simulation results include conductivity, ion diffusion coefficient and electrochemical window. The simulation results are used as performance indicators, g′=g, and the simulation results correspond to the process parameters one by one. Input the performance index corresponding to each process parameter into the trained optimization analysis model, predict the judgment label corresponding to each process parameter, obtain the corresponding judgment result according to the predicted judgment label, and judge the optimization direction of the performance index corresponding to the process parameter. The optimization direction is the peak value or valley value of the performance index during the optimization process. According to the judgment label corresponding to each process parameter, j′ simulation results are screened out from g′ simulation results corresponding to each process parameter, and the corresponding j groups of process parameters are obtained according to the j′ simulation results, and marked as excellent process parameters; wherein, j=j′, g>j>0.

6. The multi-parameter intelligent optimization method for wide temperature range lithium battery electrolyte according to claim 5, characterized in that: The training process of the optimization analysis model includes: Pre-set corresponding judgment results for multiple groups of performance indicators, the judgment results include trending peak values ​​and trending valley values, and set different digital labels for trending peak values ​​and trending valley values; Mark the digital label of the judgment result as a judgment label, and convert the performance index and the corresponding judgment label into a corresponding set of feature vectors; Each group of feature vectors is used as the input of the optimization analysis model. The optimization analysis model takes a group of prediction judgment labels corresponding to each group of performance indicators as output, and takes the actual judgment labels corresponding to each group of performance indicators as prediction targets, and the actual judgment labels are pre-set digital labels of the judgment results corresponding to the performance indicators; minimizing the sum of the prediction errors of all performance indicators is used as the training goal; the optimization analysis model is trained until the sum of the prediction errors reaches convergence and the training is stopped; the optimization analysis model is a deep neural network model.

7. The multi-parameter intelligent optimization method for wide temperature range lithium battery electrolyte according to claim 6, characterized in that: Randomly combine the optimal process parameters corresponding to each recipe parameter and the optimal recipe parameters corresponding to each process parameter into k electrolyte parameters, k = n × j, the electrolyte parameter is marked as p, p = (x, y); K electrolyte parameters are simulated to obtain the optimal electrolyte parameters as intelligent optimization parameters.

8. The multi-parameter intelligent optimization method for wide temperature range lithium battery electrolyte according to claim 7, characterized in that: The method for obtaining the optimal electrolyte parameters includes: S1: preset initialization temperature, cooling coefficient δ and maximum number of iterations And let the initialization temperature = the highest simulation temperature C MAX ; S2: Randomly set a feasible solution S, the feasible solution S is the electrolyte parameter p, the electrolyte parameter range is k electrolyte parameters, and the electrolyte parameter range is the range of the feasible solution S; S3: determine the fitness function; The fitness function is expressed as: f(x, y) = DJ, where f(x, y) is the fitness and DJ is the comprehensive score of the electrolyte parameters; S4: Calculate the fitness f(x, y) corresponding to the feasible solution S; take the feasible solution S as the current point, perform random perturbations in the neighborhood of the current point, obtain a new feasible solution S′, and calculate the fitness f(x, y)′ corresponding to the new feasible solution S′; S5: Calculate the fitness difference f(x,y)″, the expression of the fitness difference f(x,y)″ is f(x,y)″=f(x,y)′-f(x,y); If the fitness difference f(x,y)″>0, then let S=S′, that is, assign the value of the new feasible solution S′ to the feasible solution S; if the fitness difference f(x,y)″≤0, then calculate the probability k′, and let S=S′ according to the probability k′; S6: Loop S4 to S5 until the number of loops reaches the maximum number of iterations When , the loop ends and enters S7; S7: Set the current temperature to C DQ =C×δ, that is, the current temperature in S1 is cooled down, and the value after cooling is assigned to the current temperature; let the maximum number of iterations That is, assign the value of the maximum number of iterations after the reduction to the maximum number of iterations; if the maximum number of iterations after the reduction is not an integer, the maximum number of iterations after the reduction is rounded up so that the maximum number of iterations after the reduction is an integer; S8: loop S4 to S7 until the current temperature C DQ <C min When , the loop ends, the electrolyte parameters corresponding to the feasible solution S are obtained and marked as the optimal electrolyte parameters.

9. The multi-parameter intelligent optimization method for wide temperature range lithium battery electrolyte according to claim 8, characterized in that: The calculation method of the comprehensive score of the electrolyte parameters includes: Obtain conductivity, ion diffusion coefficient, and electrochemical window; After assigning preset weights to conductivity, ion diffusion coefficient, and electrochemical window, a comprehensive score of electrolyte parameters was obtained.

10. A multi-parameter intelligent optimization system for a wide temperature range lithium battery electrolyte, used to implement the multi-parameter intelligent optimization method for a wide temperature range lithium battery electrolyte according to any one of claims 1 to 9, characterized in that: include: A data acquisition module is used to collect electrolyte parameters, including formulation parameters and process parameters, and is used to collect electrolyte performance data at different temperatures; Data-driven module, used to set constraints and build data-driven models based on recipe parameters, process parameters and constraints; Process simulation module, which performs process simulation based on recipe parameters and process parameters to obtain the optimal recipe parameters and optimal process parameters; Optimization calculation module: randomly combine the optimal process parameters corresponding to each formula parameter and the optimal formula parameters corresponding to each process parameter into k electrolyte parameters, optimize and calculate the k electrolyte parameters, and obtain the optimal electrolyte parameters as intelligent optimization parameters.

Images